Trump locks Europe out of AI models; China presents the counter-proposal — a global AI platform, and everyone’s invited. What Beijing wants and why Europe remains skeptical. By founding the World Artificial Intelligence Cooperation Organization (WAICO), China has created its own international framework for governing artificial intelligence — without the players that have traditionally shaped technology policy.
For now. The founding treaty was signed on July 16, in Shanghai by representatives from 29 countries, with UN Secretary-General António Guterres in attendance. The organization is headquartered in Shanghai. Founding members include Russia, Brazil, Indonesia, Pakistan, Malaysia, Kazakhstan, Kenya, South Africa, Serbia, Belarus, Cuba, Venezuela, and Zambia, among others.
The United States, E.U. member states, and India are not participating. In Chinese diplomatic circles, WAICO is described as a “major step” through which the People’s Republic is answering the Global South’s call for greater participation in AI development. The organization is meant to serve two purposes: spreading AI applications across the Global South and organizing training and seminar programs.
The goal, Beijing says, is to use AI “for good.” Four principles — and a jab at Washington The conceptual framework rests on four principles that China’s leadership laid out at the first World AI Conference in Shanghai: openness and win-win cooperation, risk awareness, inclusivity, and solidarity in global governance.
Under win-win, Beijing groups scenario-based applications such as AI plus industry and AI plus health — in other words, integrating AI with existing economic sectors. AI must remain safe, reliable, and controllable, the argument goes. China says it is working on technical oversight, early-warning and emergency-response mechanisms, as well as laws, standards, and ethical guidelines.
Misuse is to be prevented and AI kept under human control. China rejects what it calls the overextension of national security — the principle of placing one country’s security above the interests of others. The target is clearly export controls and access restrictions.
Asked whether this language should be read as a response to the short-notice blocking of leading American models for European users, Beijing pointed out that Chinese models are open-source, rendering any accusation of bloc-building moot. The fourth principle aims at multilateralism.
Governance rules and technical standards should be coordinated internationally, and Global South countries should receive support in building their own capacities — to avoid, in China’s phrasing, new historical injustices. Over the next five years, China intends to provide 5,000 AI training slots for participants from Global South countries.
Whether these are degree programs, scholarships, or short-term courses remains unclear. Plans also include AI cooperation centers with organizations such as ASEAN, the African Union, the Arab League, BRICS, and the Shanghai Cooperation Organization. A flagship project is an AI-powered early-warning system that, according to Chinese sources, forecasts typhoons and other extreme weather events in more than 30 countries.
How China translates technological cooperation into concrete projects — across blocs and without exclusive commitments — is currently on display in Kazakhstan. In June, the Central Asian country agreed with the American firm Firebird on investments of up to ten billion dollars in a computing cluster with up to 100,000 Nvidia GPUs in the coal-mining city of Ekibastuz.